2. Agenda
Rationale: the Service Economy
Software as a Service (SaaS)
Service oriented architectures
The REST architectural style
AJAX and mashups
Cloud computing architectures
3. The Service Economy
Modern economies rely upon services
This means that several companies offer support for
activities in the primary (agricolture) or secondary
(industry of goods) sectors, or to other companies
offering services themselves
A service company is usually independent (not
owned) from other companies and their services;
however in several cases they should cooperate or at
least be coordinated in some way
4. From Products to Services
Our society is shaped by the forces of
Specialization
Standardization
Scalability
It is now mostly “service” oriented
Transportation
Telecommunication
Retail
Healthcare
Financial services
Education and training
…
5. Attributes of physical services
A service is not owned by its user (compare with product)
Well defined, easy-to-use, standardized interface
New services can be offered by combining existing services
(almost) always available but idle until requests come
“Provision-able” (used by someone only when necessary,
then reassigned to someone else)
Easily accessible and usable readily
no “integration” required
Self-contained : no visible dependencies to other services
Coarse grain
Independent of consumer context
but a service can have a context
Quantifiable quality of service
Do not compete on “What” but “How”
6. Example: mail
Customers use mail everywhere
Interfaces: PostOffice, Mail Box, stamp, postman
Apparently no relationship with other services (eg.
Transportation, payments)
Quality of service: price, delivery time, lost
messages
7. The global service economy
Top Ten Nations by Labor Force Size
(about 60% of world labor in just 10 nations)
A = Agriculture, G = Goods, S = Services
Nation % % % % 25 yr % 2008
WW A G S delta S
100%
90%
Labor 80%
70%
China 21.0 50 15 35 191 60%
50%
India 17.0 60 17 23 28 40%
30%
U.S. 4.8 3 27 70 21
20%
10% China
0%
Indonesia 3.9 45 16 39 35 1810 1835 1860 1885 1910 1935 1960 1985 2010
Brazil 3.0 23 24 53 20 100%
90% (S) Services:
Russia 2.5 12 23 65 38 80%
Value from enhancing the
capabilities of things (customizing,
70%
distributing, etc.) and interactions between things
Japan 2.4 5 25 70 40
60%
50%
40%
Nigeria 2.2 70 10 20 30 30%
20%
United States
Banglad. 2.2 63 11 26 30
10%
0%
1800 1820 1840 1860 1880 1900 1920 1940 1960 1980 2000
Germany 1.4 3 33 64 44
>50% (S) services, >33% (S) services
8. What is meant by Service?
In economics and marketing, a service is the intangible
equivalent of an economic good. Service provision is an
economic activity where the buyer does not generally, except by
exclusive contract, obtain exclusive ownership of the thing
purchased (Wikipedia, 2010)
A service is a “provider to client” interaction that creates and
captures value while sharing the risks of the interactions
Services are the application of specialized competences (skills and
knowledge) through deeds, processes, and performances for the
benefit of another entity or the entity itself (Vargo & Lusch, 2004a)
Services are value that can be rented (in the broad sense) by the
application of some process that the renter (client) participates in.
This contrasts with goods, whose value (once purchased) is owned
by the customer (Lovelock & Gummesson, 2004)
9. What is a
Service (1)
A facility supplying some
public demand
the work performed by
one that offers help, use,
or benefit
Merriam-Webster
10. What is a Service (2)
In economics and marketing, a
service is a non-material
equivalent of a good
Service composition is
managed by service providers
who have to offer or provision
more complex services
The service-good spectrum
shows that most services
include some goods, and also
that most goods involve some
service
Example: restaurant
en.wikipedia.org/wiki/Service
12. Software-as-a-Service (SaaS)
1990: Web 0.9 (Tim Berners-Lee)
1995: Web 1.0 (some dynamic content, Netscape)
1997: Services (Google, e-commerce...)
1999: LoudCLoud (first infrastructure for SaaS)
2000: Web 2.0 (rich UI's, social computing)
2004: SaaS & SOA (Google Maps, Amazon S3...)
2008: Cloud Computing (pay as you go)
12
13. Software as a service
Service: the non-material equivalent of a good
Software is a service at heart, albeit an automated one, but
it is sold much like a manufactured good. Customers have
to pay large sums of money up front, bear much of the risk
that a program may not work as promised, and cannot
readily switch vendors.
The Economist, 2003
14. Software-as-a-Service (SaaS)
Definition of SaaS: Internet-based deployment (for access
and management) of commercially available software
Service managed from “central” locations enabling
customers to access applications remotely via web
Delivery: one-to-many model (single instance, multi-tenant
architecture) including architecture, pricing, partnering, and
management characteristics
Centralized updating, which obviates the need for end-users
to download patches and upgrades
Frequent integration into a larger network of communicating
software—either as part of a mashup or a plugin to a
platform as a service
14
15. Software as a service
An operating system service?
Program execution, file management
A Software Service?
Messaging Service
Log Service
Peer to peer distribution
Search engine
A Business Service?
Customer service
Bidding service
16. Software services: a multidisciplinary problem
• Industry policies
Business • Industry imperatives
Issues • Business priorities
Business Model
Functional Issues • Functionality expressed as
Services
• Map to Business Processes
• Based on standards
Service Oriented Architecture
Operational • Model Driven Development
Issues • Middleware and Software
• Technology Platforms &
Solutions Realized on an Infrastructure Frameworks
• Network & Hardware
17. Dimensions of Sw Engineering for Services
Application Middleware Distributed
Model Computing
Mainframe MQ
Client/Server EAI, B2B DCE
Web Apps BPM CORBA, J2EE
Portals
WS, ebXML
Software
Construction
SalesForce.com
Service Orientation
SAP, Oracle… Object Orientation
MDD
Solutions XP Structured Computing
RUP Programming Model
Methodologies
18. SOC and SOA
Service Oriented Computing is a distributed
programming paradigm for business
applications supporting complex enterprises
Service Oriented Architectures are systems of
services producers and service consumers
Mostly made of independent software systems
able to be dynamically integrated
19. SOA
SOA = Service Oriented Architecture
SOA is a recent term, but in the past have been
built many systems that today we could call SOAs.
Example: the Web
The essence of a SOA lies in independent
services interconnected by messages
On the Web, a specific SOA technology is Web
Services (W3C)
21. OMG Standards for SOA
http://www.omg.org/attachments/pdf/OMG-and-the-SOA.pdf
BPMN
Business
Process UML
BPDM
SBVR
Task
definition Interfaces defined by enterprise context
Task
implementation ODM UML
Business
Services
RAS
Interfaces defined by enterprise semantics and requirements
J2EE UML
Profile
Components
CORBA
UML Profile
CWM/
Operational IMM
Resources KDM
Servers
Mainframes Servers Mainframes
Data
Data Data Data
22. Typical topics in SOA
Model, design, and implement a SOA
Create an agile, reusable SOA
Automate business processes by mapping to the
architectural model
Orchestrate services and execute processes with the
Business Process Execution Language (BPEL)
Achieve interoperability within a SOA using proven
standards and best practices
Secure and govern an enterprise SOA
23. Standardizing bodies for SOAs
W3C (1994)
OASIS (1993), consortium of former GML
providers, deals with applications using XML
OMG (1989)
WS-I (2002), promotes interoperability among the
stack of Web Services specifications"
24. Elements of SOA Design
Business Modeling
Service Oriented Architectural Modeling and Design
Model Driven assumptions (loose coupling)
Distributed objects and MOM (Message Oriented
Middleware) for component-based sw systems
27. Model driven SOA design
www.theenterprisearchitect.eu/archive/2009/06/03/a-framework-for-model-driven-soa
27
28. Services and SOAs
A service is a program interacting via message
exchanges
On Web all messages and service descriptions are
written in XML
A SOA is a set of deployed services cooperating
in a given task
Adapt to new services after deployment
29. SOA Concepts
Service manage
Operational
Requirements
bound by Expose itself using
enforce
State
Policies Edge
implement
Process/Create
governed by
Message Exchange
Pattern Messages
Contracts
describe direction & types of
contain Schemas define structure of
30. SOA Reference Architecture
AJAX Portlets WSRP B2B Other
consumers
Service Consumer
Integration (Enterprise Service Bus Approach)
Data Architecture and Business Intelligence
business processes
process choreography
Monitoring Infrastructure Service
QoS, Security, Management, and
Governance
services
atomic and composite
Service Provider
service components
operational systems Packaged Custom OO
Application Application Application
31. Main SOA Standards
BPEL (Business Process Execution Language): a standard for assembling sets
of discrete services into an end-to-end business process
J2EE: the Java Platform, Enterprise Edition, with APIs for deploying and
managing Web services
JSR 168: standard for portal and portlet interoperability
JSR 181: an API for Web services metadata annotation
SOAP (Simple Object Access Protocol): a W3C-approved standard for
exchanging information among applications
UDDI (Universal Description, Discovery, and Integration): an OASIS-approved
standard specification for defining Web service registries
WS-I (Web Services Interoperability): an open industry organization promoting
Web services interoperability across platforms, operating systems, and
languages
WS-Reliability (Web Services Reliability): a SOAP-based protocol for
exchanging SOAP messages, with delivery and message-ordering guarantees
WS-Security (Web Services Security): a SOAP-based protocol that addresses
data integrity, confidentiality, and authentication in Web services
WSDL (Web Service Description Language): a W3C-approved standard for
using XML to define Web services
WSIF (Web Services Invocation Framework): an open source standard for
specifying, in WSDL, EJB implementations for the Web server
WSRP (Web Services for Remote Portlets): an OASIS standard for integrating
remote Web services into portals
XML (Extensible Markup Language): a data markup language for Web services
32. What is SOA?
Use other services as RPC servers for your app
Web 1.0: large sites organized this way internally
Yahoo!, Amazon, Google, …: external “Services”
available, but complex: Doubleclick ads, Akamai
XML based Web Services msgs and descriptions
Web 2.0: consumer-facing service API’s
Services: Google API, Google Analytics, Amazon
CloudFront...
Platforms: Facebook, Google Maps, ...
Mashups, e.g. housingmaps.com
User-composable services, e.g. Yahoo Pipes
32
33. Amazon.com: Web 1.0 SOA
~50 “two-pizza” teams of
web server
“developer/operators” web server
web server
~10 operators
monitor the whole site
page the resolvers on alarm web server
web server
service A
web server
web server
service B
web server
web server
service C
~1000 resolvers
10-15 per team, 1 on-call 24x7
DB DB DB
monitor own service, fix problems
Over 140 code change commits/month
Internal microcosm of service-oriented architecture
(as were Yahoo, Google, others)
P. Bodík et al., Advanced Tools for Operators at Amazon.com, Proc. ICAC 2005
34. SOA = RPC
Transport: HTTP(S)
Data interchange:
XML DTD (e.g., RSS)
JSON (Javascript Object Notation)
Request protocol:
SOAP (Simple Object Access Protocol)
JSON-RPC
On the horizon: WebHooks (HTTP POST callback,
for “push”)
34
35. JSON-RPC
Open connection to designated port on server
Send HTTP method & request URI, with MIME type of body
set to application/json
Then send request body:
{ "version": "1.1",
"method": "confirmFruitPurchase",
"id": "194521489",
"params": [ [ "apple", "orange", "pear" ], 1.123 ]
}
Response might be something like this:
{ "version": "1.1",
"result": "done",
"error": null,
"id": "194521489"
}
You have to handle substantially all errors
35
36. RSS
Request is a regular HTTP GET to a specified URL
<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1">
<channel>
<title>Altarena Playhouse Ticket Availability</title>
<link>http://www.audience1st.com/altarena/store</link>
<description>Altarena Ticket Availability</description>
<item>
<title>Sylvia - Friday, May 14, 8:00 PM – Buy now</title>
<link>http://.../store?showdate_id=347</link>
<guid permalink="false">http://.../store?ts=1271058414</
guid>
</item>
<item>
...
</item>
</channel>
</rss> 36
37. AJAX and SOA
AJAX: client server
client makes (async) requests to HTTP server
client-side JavaScript upcall receives reply and
decides what to do
commonly, response includes XHTML/XML to update
page, or JavaScript to execute
Doesn’t really make sense except in context of client
SOA: server server or client server
one principal makes (sync or async) requests to an
HTTP server
formerly, principal was a server running some app
today, powerful JavaScript clients blur the line
37
38. Two ways to do it...
“Thin”
Web 2.0
browser Craigslist.org
app
client
“Fat” browser
Google Maps
client
+ Client portability
+/– Client performance (both app download &
JavaScript execution)
+ Availability of utility libraries for app development
– Privacy/trustworthiness of aggregator app
– Caching
38
39. REST (Representational State Transfer) style
Architectural style:
Client-server, stateless, cache
description of properties that made Web 1.0 successful
by constraining SOA interactions
In context of SOA for Web 2.0
HTTP is transport; HTTP methods (get, put, etc.) are the
only commands
URI names are a resource
Client has resource has enough info to request
modification of the resource on server
A cookie can encode part of transferred state
If your app is RESTful, it’s easy to “SOA”-ify
39
40. What is REST?
REST is a term coined by Roy Fielding in his Ph.D
dissertation to describe an architectural style of
networked software systems
"Representation State Transfer is intended to
evoke an image of how a well-designed Web
application behaves: a network of web pages (a
virtual state-machine), where the user progresses
through an application by selecting links (state
transitions), resulting in the next page (representing
the next state of the application) being transferred
to the user and rendered for use."
http://www.ebuilt.com/fielding/pubs/dissertation/top.htm
41. Why is it called
"Representation State Transfer"?
http://www.boeing.com/aircraft/747
Client Resource
Fuel requirements
Maintenance
...
747.html
The Client references a Web resource using a URL. A representation of the resource
is returned (in this case as an HTML document).
The representation (e.g., Boeing747.html) places the client application in a state.
The result of the client traversing a hyperlink 747.html is another resource is accessed.
The new representation places the client application into yet another state.
Thus, the client application changes (transfers)
state with each resource representation --> Representation State Transfer!
42. Motivation for REST
"The motivation for developing REST was to create
an architectural model for how the Web should work,
such that it could serve as a framework for the Web protocol standards.
REST has been applied to describe the desired Web architecture,
help identify existing problems, compare alternative solutions,
and ensure that protocol extensions
would not violate the core constraints that make the Web successful."
- Roy Fielding
43. REST with HTTP examples
HTTP GET! HTTP PUT! HTTP POST! HTTP DELETE!
Collection URI, such as http://example.com/customers/257/orders
List the members of Replace the Create a new entry in the delete the
the collection, entire collection collection. The ID created is entire
complete with their with another usually included as part of collection"
member URIs for collection" the data returned by this
further navigation" operation."
HTTP GET! HTTP PUT! HTTP POST! HTTP DELETE!
Element URI, such as http://example.com/resources/7HOU57Y
Retrieve a Update (or Treats the addressed Delete the
representation of create) the member as a addressed member
the addressed addressed collection in its own of the collection. ""
member of the member of the right and creates a "
collection in an collection" new subordinate of it."
appropriate MIME
type"
43
44. Example: REST vs SOAP-based
A company deploying 3 Web services to enable its
customers to:
get a list of parts
get detailed information about a particular part
submit a Purchase Order (PO)
See the REST solution first, then the SOAP
solution
44
45. The REST way of Implementing the
Web Services
HTTP GET request URL 1
Parts
Response List
(HTML/XML doc) HTTP response
Web Server
HTTP GET request URL 2
Part
Response
(HTML/XML doc) HTTP response
HTTP POST (HTML/XML)
PO
URL 3
PO
URL to submitted PO HTTP response
46. Implementing the Web
Service using SOAP
• Service: Get detailed information about a particular part
– The client creates a SOAP document that specifies the procedure
desired, along with the part-id parameter.
<?xml version="1.0"?>
<soap:Envelope xmlns:soap="http://schemas.xmlsoap.org/soap/envelope/">
<soap:Body>
<p:getPart xmlns:p="http://www.parts-depot.com">
<part-id>00345</part-id>
</p:getPart>
</soap:Body>
</soap:Envelope>
Again, the client will HTTP POST this document to the SOAP server at:
http://www.parts-depot.com/soap/servlet/messagerouter
Note that this is the same URL as was used when requesting the parts list.
The SOAP server peeks into this document to determine what procedure to invoke.
47. Implementing the Web Services
using SOAP
SOAP envelope
Request
HTTP POST
(XML doc) URL 1
getPartsList()
Response
HTTP Response
(XML doc)
Request
Web Server
HTTP POST
(XML doc) URL 1
Response
SOAP Server getPart(id)
HTTP Response
(XML doc)
PO
HTTP POST
(XML doc) URL 1
submit(PO)
Response
HTTP Response
(XML doc)
Note the use of the same URL (URL 1) for all transactions.
The SOAP Server parses the SOAP message to determine which method to invoke.
All SOAP messages are sent using an HTTP POST.
49. XML (eXtensible Markup Language)
<?xml version="1.0" encoding="UTF-8"?>
<book year="1967">
<title>The politics of experience</title>
<author>
<firstname>Ronald</firstname>
<lastname>Laing</lastname>
</author>
</book>
Really a metalanguage for describing
hierarchical, semistructured, schema-less data
XML Document Type Definition (DTD) specifies
structural & content constraints on a particular
document type
49
50. XML (eXtensible Markup Language)
<?xml version="1.0" encoding="UTF-8"?>
<book year="1967"> Attribute Value
<title>The politics of experience</title>
<author> Element
<firstname>Ronald</firstname> Element
<lastname>Laing</lastname>
</author>
</book>
Really a metalanguage for describing
hierarchical, semistructured, schema-less data
XML Document Type Definition (DTD) specifies
structural & content constraints on a particular
document type
50
51. XHTML & CSS
XHTML: a document conforming to a particular DTD
describing a hierarchical collection of HTML elements
Variants: Strict, loose, transitional (for compatibility with deterioriating HTML
syntax 1990-95)
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
inline (headings, tables, lists...)
embedded (images, video, Java applets, JavaScript code...)
fill-in forms—text, radio/check buttons, dropdown menus...,
marshaling arguments into either URI or request body
CSS (Cascading Stylesheets) for presentation
Strict XHML forbids presentational markup
Idea: complete separation of appearance from structure
51
52. CSS Styles get visual styling applying selectors
<link rel="stylesheet" href="mystyles.css"/>
<div class="pageFrame" id="pageHead">
<h1>
Welcome,
<span id="custName">Armando</span>
</h1>
</div>
In mystyles.css (static asset with MIME type text/css):
div.pageFrame { background-image: url('/banner.gif'); }
h1 { font-size: large; float: left; }
#custName:hover { background-color: yellow; font-weight: bold; }
Style properties include borders, background images, and layout
directives (floating, absolute positioning, min/max scaled sizes, etc.)
Changing style properties has side effect of re-rendering
e.g., change display or visibility property to show/hide elements
55. Dynamic content generation
Most Web 1.0 (e-commerce) sites actually run a
program that generates the output
Originally: templates with embedded code
“snippets”
Eventually, embedded code became “tail that
wagged the dog” and moved out of the Web server
Languages/frameworks evolved to capture
common tasks
Perl, PHP, Python, Ruby on Rails, ASP, ASP.NET, Java
Servlet Pages, Java Beans/J2EE, ...
55
56. SaaS 3-tiers architecture
Common gateway interface (cgi): allows a Web persistent
storage
server to run a program storage
Server maps some URI’s to application names
When the app is run, it gets the complete HTTP
request including headers app
application
server
“Arguments” embedded in URL with “&”
syntax or sent as request body (with POST)
http://www.foo.com/search?term=white
%20rabbit&show=10&page=1 HTTP
App generates the entire response server
content (HTML? an image? some javascript?)
HTTP headers & response code
Plug-in modules for Web servers allow long-
running CGI programs & link to language
interpreters
• Various frameworks have evolved to capture this structure!
56
57. SaaS 3-tiers deployment
HTTP server (“web server”)
persistent
“fat” (e.g. Apache): support virtual hosts, plugins for storage
storage
multiple languages, URL rewriting, reverse proxying, ....
“thin” (nginx, thin, Tomcat, ...): bare-bones machinery to
support one language/framework; no frills app
application
Application server server
1. separate server process, front-ended by a “thin” HTTP
server
2. or linked to an Apache worker via FastCGI or web server HTTP
plug-in: mod_perl, mod_php, mod_rails, ... server
Apache can spawn/quiesce/reap independent processes
Persistent storage
Common RDBMS (MySQL, PostgreSQL, etc.)
communicate w/app via proprietary or standardized
database “connector” (ODBC, JDBC, ...)
Hence LAMP: Linux, Apache, MySQL, PHP/Perl
57
58. Frameworks
Support for more languages: Apache modules
(mod_perl, mod_php, mod_rails ...)
avoid spawning new process per request
typically embed language interpreter in Apache
Support for common idioms like sessions
Cookie management
virtualize connection to database
“dispatcher” interactions with front-end HTTP server
Early “templating systems” (e.g. PHP) vs. modern
“full stack frameworks” (e.g. Rails)
58
59. Example: Rails, a Ruby-based MVC Framework
• Implemented almost entirely in Ruby
• Distributed as a Ruby “gem” (collection of related libraries & tools)
• Connectors for most popular databases
Relational mysql or
Model, View, Controller tables Database
sqlite3
Subclasses of
ActiveRecord::Base models/*.rb Ruby
your app interpreter
views/*.html.erb controllers/*.rb
Rails Rails CGI or other dispatching
rendering routing
apache
Subclasses of Subclasses of
ActionView ApplicationController firefox
60. What’s new in Web 2.0?
Primitive UI => Rich UI
enable “desktop-like” interactive Web apps
enable browser as universal app platform on cell phones
“Mass customize” to consumer => Social computing
tagging (Digg), collaborative filtering (Amazon reviews), etc. =>
primary value from users & their social networks
write-heavy workloads (Web 1.0 was read-mostly)
lots of short writes with hard-to-capture locality (hard to shard)
Libraries => Service-oriented architecture
Integrate power of other sites with your own (e.g. mashups that
exploit Google Maps; Google Checkout shopping cart/payment)
Pay-as-you-go democratization of “services are king”
Focus on your core innovation
Buy & rack => Pay-as-you-go Cloud Computing
60
61. Rich Internet Apps (RIAs)
Closing gap between desktop & Web
Highly responsive UI’s that don’t require server roundtrip per-action
More flexible drawing/rendering facilities (e.g. sprite-based animation)
Implies sophisticated client-side programmability
Local storage, so can function when disconnected
early example: Google Docs + Google Gears
include offline support, local storage, support for video, support for arbitrary
drawing, ...
currently many technologies—Google Gears, Flash, MS
Silverlight...
client interpreter must be embedded in browser (plugin, extension, etc.)
typically has access to low-level browser state => new security issues
N choices for framework * M browsers = N*M security headaches
proposed HTML5 may obsolete some of these
61
62. Rich UI with AJAX
(Asynchronous Javascript and XML)
Web 1.0 GUI: click → page reload
Web 2.0: click → page can update in place
also timer-based interactions, drag-and-drop,
animations, etc.
How is this done?
1. Document Object Model (1998, W3C) represents
document as a hierarchy of elements
2. JavaScript (1995; now ECMAScript) makes DOM
available programmatically, allowing modification of
page elements after page loaded
3. XMLHttpRequest (2000) allows async HTTP
transactions decoupled from page reload
4. JavaScript libraries (jQuery, Prototype,
script.aculo.us) encapsulate useful abstractions
62
63. DOM & JavaScript:
Document = tree of objects
hierarchical object model representing HTML or
XML doc
Exposed to JavaScript interpreter
Inspect DOM element value/attribs
Change value/attribs → redisplay or fetch new
content from server
Every element can be given a unique ID
JavaScript code can walk the DOM tree or select
specific nodes via provided methods
<input type="text" name="phone_number" id="phone_number"/>!
<script type="text/javascript">!
var phone = document.getElementById('phone_number');!
phone.value='555-1212';!
phone.disabled=true;!
document.images[0].src="http://.../some_other_image.jpg";!
</script>!
63
64. JavaScript
A browser-embedded scripting language
OOP: classes, objects, first-class functions, closures
dynamic: dynamic types, code generation at runtime
JS code can be embedded inline into document...
<script type="text/javascript"> <!-- # protect older browsers
calculate = function() { ... } // -->
</script>
...or referenced remotely: <script src="http://evil.com/Pwn.js"/>
Current page DOM available via window, document objects
Handlers (callbacks) for UI & timer events can be attached to JS
code, either inline or by function name: onClick, onMouseOver,...
Changing attributes/values of DOM elements has side-effects, e.g.:
<a href="#" onClick="this.innerHTML='Presto!'">Click me</a>
64
65. AJAX = Asynchronous Javascript and Xml
Recipe:
attach JS handlers to events on DOM
objects
in handler, inspect/modify DOM elements
and optionally do asynchronous HTTP
request to server
register callback to receive server response
response callback modified DOM using
server-provided info
JavaScript as a target language
Google Web Toolkit (GWT): compile Java => emit JS
Rails: runtime code generation ties app abstractions to JS
65
66. A browser is an application container
“Web apps” include animation, sound, 3D graphics,
disconnection, responsive GUI...
Browser =~ new OS: manage mutually-untrusting apps (sites)
Source: Robert OʼCallahan (Mozilla.org),!
Inside Firefox!
66
67. Social Computing
Web 1.0: add value via mass customization
select content/presentation for you based on best guesses about
your interests
resource: demographic/analytic data about you
Web 2.0: add value via connecting to social network
vendor: your friends’ interests are a good indicator of your interests
user: value added to existing content == how your friends interact
with it
resource: your social network
From social networking site to social network as a way of
structuring applications
67
68. Social Computing
Amount of content “created” by each user small!
e.g., Digg article, rate video, play a Facebook game
but still creates lots of short random writes
consider “Like” feature on Facebook
social graphs naturally hard to partition (though would
love to see a paper about this from FB)
question for Web 2.0 developers is not whether
social computing is part of your app, but how
later we will discuss technical architecture of
“connecting” an app to social networks
68
69. Facebook
Facebook plug-in apps
FBQL!
FB data!
1.
Facebook.com! 2.! Your app!
html 4. 3.!
AJAX SO
• Facebook platform (“Facebook Connect”)!
A
FB data!
REST!
1.
html+ Your app! 2 (opt.). Facebook.com!
3.
xfbml
4.
èREST via JavaScript &
XFBML!
69
ç HTML IFRAME w/FB content!
70. Google Maps
Your app embeds Javascript-heavy client code (provided
by Google)
client-side functionality: clear/draw overlays, etc.
server-side functionality: fetch new map, rescale, geocoding
Attach callbacks (handled by your app) to UI actions
Result of callback can trigger additional calls to Google
Maps code, which in turn contact GMaps servers
1. Your Google
2. app Maps
html+
js 4.
3.
70
72. Scaling via Replication
The “most general” deployment scenario for
a 3 tier Web application
DB DB?
Many Web servers
possibly including static-asset servers
L4/L7 load balancers distribute load among them App … App
Caches and reverse proxies remember
previously-computed content $ … $
whole page caching
page fragment caching, query caching WS … WS
Apache in reverse-proxy mode, or
memcached process(es) addressed by Asset
app server LB …LB
Svr
Integration of caching with application logic the Internets
varies by framework
72
73. “Scale makes availability affordable”
Goal: interchangeability (send any
user request to any available server) DB DB?
each server handles 1/N load
affinity can be used to “soft-pin” users to App … App
particular servers
requires good support for session state $ $
…
abstraction in app framework
lose 1 server => lose 1/N capacity WS … WS
Load Balancers have logic to detect
failed servers & remove from rotation Asset LB …LB
Svr
until they are resurrected
the Internets
73
74. Asset Servers
For serving static assets (images, sound clips,
CSS, etc.)
Separate Web server process, configuration
optimized for fast static file serving
Web 2.0: use Amazon S3 (blob store) or
CloudFront (CDN)
helps to have good asset-server abstraction in app
framework
74
75. Deployment scenarios (& approximate pricing)
Buy/rack/install/configure it yourself...that’s so Web 1.0
Shared hosting ($3/month)
turnkey support for popular frameworks, hosted versions of popular
building blocks (e.g. MySQL)
highly variable performance, multitenant per machine
Virtual private host ($10/month)
better isolation and security through virtualization
substantially more administration
“Framework VM” or “curated” environments (Heroku,
Google AppEngine, Force.com) – pricing varies
hosted extensions: memcached, profiling, etc.
integration of 3rd party hosted services, e.g. Amazon S3 backup
Cloud Computing
75
77. Cloud Economics: user side
Cloud Computing: User side "
Static provisioning for peak"
- wasteful, but necessary for SLA"
Capacity $
Machines
Demand Capacity
Demand
Time Time
“Statically provisioned” “Virtual” data center
data center in the cloud
Unused resources
77
78. Cloud Economics 101
Cloud Computing Provider: Could save energy
"
Capacity
Energy
Machines
Demand Capacity
Demand
Time Time
“Statically provisioned” Real data center
data center in the cloud
Unused resources
78
80. “Risk Transfer” to Cloud
“Cost associativity” from linear pricing: 1,000 CPUs for 1
hour same price as 1 CPUs for 1,000 hours (@$0.10/hour)
Washington Post converted Hillary Clinton’s travel documents to
post on WWW <1 day after released
Lab graduate students demonstrate improved Hadoop (batch job)
scheduler—on 1,000 servers
Major enabler for SaaS startups
Animoto traffic doubled every 12 hours for 3 days when released as
Facebook plug-in
Scaled from 50 to >3500 servers
...then scaled back down
Goal: fix any transient problem by adding/removing nodes
Single-node performance becomes much less important
80
81. Classifying Clouds for Web 2.0
Instruction Set VM (Amazon EC2)
Managed runtime VM (Microsoft Azure)
Curated “IDE-as-a-service” (Heroku)
Platform as service (Google AppEngine, Force.com)
flexibility/portability vs. built-in functionality
Higher-level,
Lower-level,
More managed,
Less managed
more value-added SW
EC2 Joyent Azure Heroku, Force.com
AppEngine
81
82. Deployment
“Deployment-as-a-service” increasingly common
monthly pay-as-you-go curated environment (Heroku)
hourly pay-as-you-go cloud computing (EC2)
hybrid: overflow from fixed capacity to elastic capacity
Remember administration costs when comparing!
Good framework can help at deployment time
Separate abstractions for different types of state: session state,
asset server, caching, database
ORM – natural fit for social computing, and abstracts away from
SQL (vs Web 1.0 PHP, e.g.)
REST – encourages you to make your app RESTful from start, so
that “SOA”-ifying it is trivial
Scaling structured storage: open challenge
82
83. Cloud Computing
CC makes it possible to obtain results on 1000’s of
servers
Es. BOOM (declarative cloud programming)
boom.cs.berkeley.edu
Eucalyptus www.eucalyptus.com makes hybrid
cloud computing reasonably practical
Run small experiments locally, then “scale up” to cloud
for results
Why aren’t you using cloud computing yet?
83
84. Example: Facebook
Facebook has 2 datacenters, 1 per coast
reads spread across both
writes only to W. Coast; periodically (~10 minutes)
replicated to E. Coast
>2000 MySQL servers, >25TB RAM for memcached
Challenge: inconsistency due to stale data
I change status message => Friends on East Coast
datacenter do not see my change for 10 min
What if E.Coast person changes his own status??
84
86. SCADS: Scalable, Consistency-Adjustable Data Storage
Most popular websites follow the same pattern
Outgrow initial prototype (on MySQL) due to scale
Build large, complicated ad-hoc systems to deal with
scaling limitations as they arise
Want Scale Independence as new users join:
No changes to application
Cost per user & request latency don’t increase
Key Innovations
1. Performance-{safe,insightful} query language
2. Declarative performance/consistency tradeoffs
3. Automatic scale up & down using machine learning
M. Armbrust et al., SCADS: Scalable Consistency-Adjustable Data Storage for Interactive Applications. Proc. CIDR 2009!
M. Armbrust et al., PIQL: A Performance-Insightful Query Language. Proc. SOCC 2010. !
87. Understanding workload spikes
death of Michael Jackson: 23% of tweets, 15% of Wikipedia
traffic...how to understand/synthesize spikes?
characterized 5 real spikes
large differences in many characteristics: steepness, magnitude, #
data hotspots, …
spike synthesis tool generates realistic workload with few
parameters using statistically sound model
Cloud computing offers alternative to overprovisioning?
model
Explora/on
of
performance
[ACDC
’09]
policy
Policy
simulator
[HotCloud
’09]
workload
ac/ons
statefu Resource
alloca/on
for
stateful
l
systems
system
P. Bodik et al., Characterizing, modeling, and generating workload spikes for stateful services, SOCC 2010!
P. Bodik et al., Statistical machine learning makes automatic control practical for Internet datacenters, HotCloud '09!
P. Bodik et al., Automatic exploration of datacenter performance regimes, ACDC '09 !
!
!
88. Browsers are more and more fat
Performance, security and GUI choke point
Cookie management, JavaScript, XSS attacks
CSS layout, side-effects of DOM manipulation
AJAX: document can change anytime after loading
RIA frameworks/plugins: Flash, HTML 5, Silverlight
Device variation: new twist due to power limits
See http://www.google.com/googlebooks/chrome for a
great fun-to-read description in comic-book form
Bring “desktop-quality” browsing to mobile tel
Enabled by 4G networks, better output devices
Parsing, rendering, scripting are all potential bottlenecks
88
89. Parallelizing the Browser (R.Bodík et al.)
New algorithms for CSS, layout, rendering, up to
80X speedup on popular web sites
Constraint-based browser programming language
unifies layout, rendering, scripting
Slashdot (CSS Selectors)
50
45
40 2ms
Speedup 35
30
25
20
15
10
5
0 84ms
1 2 3 4 5 6 7 8
Hardware Contexts
89
90. Try it yourself
1. Download & install RoR (rubyonrails.org)
A good intro book: Simply Rails 2 by Patrick Lenz
2. Signup for a free Heroku.com trial account
3. Install various useful “Ruby gems” (libraries):
gem install heroku cucumber-rails rspec-rails
4. Create app skeleton:
rails mynewapp ; git commit –m “new app” mynewapp
heroku create mynewapp
5. Start writing user stories and track your progress
(PivotalTracker.com)
6. When ready to deploy on Heroku:
git push heroku master
More resources: http://radlab.cs.berkeley.edu > Courses
90
91. Summary
New ecosystem: changes software process,
deployment process, interdependencies of DADO
(everything-as-a-service)
Web 2.0 tools have reached a new plateau of
productivity
Gap closing between “web app” user experience
vs. desktop app
New interesting research problems
91
92. Self test
Which are the consequences of defining software a
service insted of a good?
What are the main original features of the REST
style?
Which software architectural issues are typical of
social software?
What are the economic basis of cloud computing?
92
